A Smart Hydration Device for Children: Leveraging TRIZ Methodology to Combat Dehydration and Enhance Cognitive Performance

Abstract

Amid globalization and rising global temperatures, dehydration has emerged as a critical issue, especially for children who are more vulnerable due to their higher body surface-to-weight ratio. The issue is even more concerning given that adequate water intake is important for cognitive development, particularly in children since brain development is critical during early years. This study addressed this challenge by, first, designing a smart hydration device based on the Theory of Inventive Problem Solving (TRIZ). Then, this study proceeded with prototyping and testing the smart hydration device to promote increased daily water intake among Malaysian children. The device demonstrated improved water consumption and increased drinking frequency among children. Additionally, the children displayed improved cognitive performance. However, this study was limited to a specific age group and the device requires adult supervision for charging. Therefore, further research is necessary to tackle these limitations. Nevertheless, this smart device represents a promising step forward in fostering better hydration habits among children.

1. Introduction

Heat is a critical and growing threat to human health, and as global temperatures rise, instances of heat stress can become increasingly common [1]. Heat naturally flows from hotter areas to cooler ones. Thus, in warm environments where air temperature exceeds skin temperature (the commonly used reference value is 35 °C [2]), heat transfers into the body. The body’s primary cooling mechanism is sweating, which allows for heat dissipation from the body. However, in humid environments where the air is already saturated with water vapor, it reduces the rate of sweating [3]. As such, the body fails to regulate temperature, leading to the build-up of heat, known as heat stress. Malaysia, situated in Southeast Asia, has a high humidity level (ranging from 75% to 95% relative humidity) [4] and rising temperatures (0.24 °C per decade in Peninsular Malaysia from 1979 to 2023) [5]. Figure 1 displays a map of all-time record high temperatures in Malaysia, which range between 36 °C and 46 °C [6]. This highlights the increasing risk of heat-related illnesses among Malaysians [1].

During the past 2015–2016 El Nino cycle, Malaysia reported 200 cases of heat-related illnesses, out of which there were two fatalities. Meanwhile, the recent 2023–2024 El Nino occurrence proved to be more deadly with five fatalities out of the 127 reported cases of heat-related illnesses. Out of these five fatalities, three were among children, recorded as under the age of 5 [1]. Due to their higher body surface-to-weight ratio, children are at greater risk of heat-related illness compared to adults [7]. This highlights the vulnerability children face in managing heat stress compared to adults.

To remediate this, children need to maintain adequate hydration to help regulate their body temperature and reduce the risk of heat-related illnesses. The World Health Organization (WHO) defined children as those aged between 1 and 12 years, with the further subcategories of young child (1 to 4 years), toddler (2 to 3 years), and older child (5 to 12 years) [8]. On the other hand, the European Food Safety Authority (EFSA) Panel on Dietetic Products, Nutrition, and Allergies (NDA) provided guidelines on the adequate total water intake according to several age ranges [9], as displayed in

Table 1. The guidelines for adequate water intake according to age range [9].

Age Range (Years)	Adequate Intake (AI) (mL/day)
	Boy	Girl
1–2	1100 to 1200
2–3	1300
4–8	1600
9–13	2100	1900
14 and above	2500	2000

. These adequate intakes apply only to conditions of moderate environmental temperatures and physical activity levels.

However, in a study conducted on 207 children, aged between 7 and 13 years, in Muar Johor, Malaysia, it was found that the total fluid intake for boys was 1337 mL/day while for girls it was 1282 mL/day [10]. This highlights the ~30% below recommended fluid intake among Malaysian children, an alarmingly dangerous issue, exacerbated by the country’s increasing temperatures and high humidity.

Beyond its relation with fatalities, dehydration has also been reported to impair cognitive performance [11,12,13]. Cognitive performance refers to the performance of the human brain to process, store, and extract information, including processes such as attention, memory, and reasoning ability. Thus, cognitive ability is often associated with academic performance. Through the analysis of four models of comprehensive academic performance, Chinese academic performance, mathematics academic performance, and English academic performance, it was showed that cognitive ability has a significant effect on academic performance [14]. Therefore, it is only natural to emphasize the importance of cognitive development in children, as supported by various studies published on the topic [15,16,17,18,19,20,21]. This is because healthy cognitive development during childhood and adolescence is of paramount importance as it lays the foundation for a lifetime of learning, problem-solving, and decision-making [22,23].

Various initiatives have been implemented by the Malaysian Government to encourage children to drink more water. One of them is the Interactive Malaysian Childhood Healthy Lifestyle (i-MaCHeL) programme Module 13: Refresh yourself with plain water, where parents monitor their children’s water intake using a sticker book. Parents are tasked to explain the need of drinking plenty of plain water for children’s health [24]. Additionally, the Malaysian Dietary Guidelines for Children, Key Message 13: Drink plenty of water daily, also suggests ways to achieve adequate amounts of water intake daily. This includes ensuring plain water is available at all times, both at school and at home, ensuring the child drinks more water when he is active, and giving more water when your child is sick [23]. There is also an attempt in Malaysia to break the record for the most people drinking water simultaneously, in hopes of instilling healthy habits in students [25]. Nevertheless, the total fluid intake for children in Malaysia is still relatively below the recommended volume, as previously mentioned.

Additionally, the role of parents in the water intake among children has been reported to be significant. This is because children sometimes do not even know that they are thirsty. Consuming water comes to their mind when they see the surroundings or when they see water in their parents’ room or kitchen. In this regard, it has been observed that children acquire drinking behavior in the manner of modeling. This is also known as the immature thirst [13]. Similarly, the same observation has been reported where parents’ water intake is positively associated with their children’s water intake. This suggests that parents’ hydration habits affect their children [26]. As such, a parent–child interaction solution may be a good direction for instilling hydration habits in children.

TRIZ refers to the Theory of Inventive Problem Solving, invented by Genrich Saulovich Altshuller. It is one of the most comprehensive systematic innovation and creativity methodologies available to mankind [27]. To develop TRIZ, the inventor and his associates studied a vast range of technological solutions, patents, and inventions, and they extracted a number of common solution patterns which existed among them [28,29]. Now, modern TRIZ is a combination of a theory of inventive problem solving and systems evolution, analytical tools and methods for problem-solving and analysis, collections of patterns of solutions, databases of specific effects and technologies, and techniques for creative imagination and development [30]. The effectiveness of TRIZ as a problem-solving tool is validated by numerous studies that have successfully applied it across a wide range of challenges [31,32,33,34,35,36]. Thus, TRIZ can be a great tool in developing device-oriented solutions to improve Malaysian children’s water-drinking habits.

Most research emphasizes promoting hydration practices [37,38,39] rather than creating tools to encourage these behaviors. Moreover, there is also a notable gap in parent–child interactive device solutions for the same issue. To encourage Malaysian children to drink more water, necessary steps need to be taken to address this gap. While such device-oriented concepts have been proposed, that study focused solely on conceptual ideas and lacked prototype-based validation [40]. Building on that work, this study aims to design a smart system which incorporates parent–child interaction, based on TRIZ, to encourage children to increase their daily water intake.

2. Review of Existing Solutions and Research

Several studies have partially addressed this issue. For instance, Tate et al. (2012) found that replacing caloric beverages with non-caloric options, including water, led to an improved hydration level. The replacement was conducted on overweight and obese adults over a period of 6 months [41]. Similarly, Fadda et al. (2012) demonstrated that providing supplementary water during school hours positively impacted children’s transient states, such as reduced fatigue and enhanced cognitive performance. The investigation was conducted on 168 children, aged between 9 and 11 years, who were living in a hot climate (South Italy, Sardinia) [42]. Edmonds and Burford (2009) studied the effects of drinking water on cognition among 58 children, aged between 7 and 9. The children were divided into two groups, one receiving additional water and the other not. The group with extra water reported feeling less thirsty and demonstrated improved visual attention and task performance compared to the control group [43]. Studies by Tate et al. (2012), Fadda et al. (2012), and Edmonds and Burford (2009) all highlight the benefits of providing water to children, rather than relying on cultivating the habit of drinking water.

To address this issue, several devices have been developed, though most target individuals rather than children specifically. Examples include:

• The Communication Water Bottle [44];
• Dieter’s Water Intake Quantity Tracking Vessel [45];
• Liquid Consumption Counters [46];
• Monitoring Water Drinking Device [47];
• Drinking Water Reminding System [48].

These devices were designed to remind individuals to drink water and maintain daily hydration. However, they primarily rely on independent monitoring by the user. Currently, there is no device that not only reminds children to drink water but also incorporates interactive and educational elements involving other parties, such as parents, siblings, or peers. This study aims to propose device-oriented solutions that enhance children’s water-drinking behavior while providing an interactive experience.

Building on the findings by Tan, Ng, Tan, and Lim (2021) [40], and drawing from prior research on TRIZ methodologies [49,50,51] and ergonomics [52,53,54,55,56], two concepts were proposed: a dual-compartment water bottle and a measurement, accumulation, and display device.

3. Methodology

The initial stage of analysis, which built the foundation of this study, is already documented in previously published work [40]. Figure 2 displays the TRIZ process flow of the work. The analysis involves 4 steps, with Step 1 being the definition of the main problem, which was defined as “Children are not drinking enough water”. Then, Step 2 is the cause-and-effect chain analysis, where the question “why” is asked with regard to the occurrence of the main problem until potential root causes are deduced at the end of the chain. Three possible root causes were determined, one of which was “The lack of devices that aid in training children to cultivate a water-drinking habit”. The same cause-and-effect chain analysis has been previously reported in improving the quality of canned food production [57] and altering the surface chemicals and physical properties of substrates in plasma modification [58], showcasing its benefit in potential root causes identification.

Following that is Step 3, which is the Engineering Contradiction, involving the TRIZ Contradiction Matrix and 40 Inventive Principles. Using the root cause above, the technical contradiction formulated was “IF teachers/parents possess tools to train children in drinking more water, THEN children will have more opportunities to practise and eventually cultivate a water-drinking habit, BUT it will take a longer time to effectively train children”. Based on this technical contradiction, the respective improving variable (#38 Extent of automation) and worsening variables (#15 Duration of action of moving object and #25 Loss of time) were assigned, where each system pair produces a list of inventive principles, as shown in Figure 2. Out of the six inventive principles listed for the first root cause, #6 Universality was selected due to the aim that the device be not only a training tool for children, but also help them cultivate a water-drinking habit. Similarly, the same process was conducted for the other two root causes, which led to the selection of inventive principles #10 Preliminary/Prior Action and #1 Segmentation.

In a reported study, similar inventive principles of #6 Universality and #10 Preliminary/Prior Action were utilized during the redesigning of a Printed Circuit Board (PCB) casing, which significantly reduced the number of parts from 18 to 2, highlighting the successful application of said inventive principles [59]. On the other hand, the application of inventive principle #1 Segmentation can be clearly observed in the maritime industry, where the contradiction “increasing the area of a sail to capture more wind leads to decreasing controllability over the sail” was solved by splitting one large sail into a number of smaller sails [60].

Finally, Step 4 is the Idea Generation, where the preliminary idea “Dual-compartment water bottle” was derived from #1 Segmentation, while the preliminary idea “Measurement, accumulation, and display device” was derived from #6 Universality and #10 Preliminary/Prior Action. Idea generation through TRIZ has been reported to be an innovative pattern-based technique that provides designers with guidelines to support and reflect design strategies and achieve the aims of creative problem-solving and novel idea generation [61]. As such, this study is built on the idea generation by TRIZ previously published [40]. This study begins with providing the characteristics of each preliminary design, as shown in

Table 2. Preliminary ideas obtained from TRIZ methodology.

Preliminary Ideas	Characteristics/Advantages of Preliminary Idea	Possible Limitations Observed
Preliminary Idea 1: Dual Compartment Water Bottle	Constructed from the TRIZ inventive principle of segmentation. The water bottle has two compartments. The top compartment is filled with water. The bottom compartment is filled with any of the child’s favorite beverages (as a reward). The mechanism is adapted from a simple rubber screw stopper, whereby rotating the bottom compartment will loosen the screw to enable the reward beverage to flow in the top compartment. This idea is suitable for children (simple).	Some studies showed that extrinsic rewards will affect a person’s intrinsic motivation [32]. It is possible the children will remain dependent on rewards and not develop good drinking water habits.
Preliminary Idea 2: Water Measuring, Accumulating, and Displaying (M.A.D) Device Version 1	Combination of the first and second root causes. The inventive principles used are universality and preliminary action/prior action. The idea fulfils the principle of universality by being a combination of three functions: ○ Water measuring; ○ Water accumulating; ○ Water displaying. Adapting the most direct approach by finding a method of remembering or recording the amount of water drunk by the child. Water is placed on the spring which will rotate the shaft, which will push the indicator to rise accordingly to indicate the amount of water placed on the spring.	Mechanically this idea is incomplete. However, the concept is present. Hence, the concept was preserved while the mechanical concept of this device was further improved to produce the next preliminary idea.
Preliminary Idea 3: Water Measuring, Accumulating, and Displaying (M.A.D) Device Version 2	Has the same concept and functions as in M.A.D version 1 but improved mechanical design. Circular platform is suspended with an extension spring and a rack is attached by the side. When a water bottle is placed on the platform, the spring will extend and lower the rack that will turn the main shaft. The user can select between three secondary shafts by sliding just one and engaging the friction pads. The selected shaft will rotate accordingly, thus displaying the amount of water.	The spring system where the platform is constantly suspended on a spring. Base would be unstable which may affect the performance of the device. The friction between pads might not be sufficient. Accuracy will be affected if slipping of friction pads occurs.
Preliminary Idea 4: Water Measuring, Accumulating, and Displaying (M.A.D) Device Version 3	The base of device is designed to be much wider. Friction pads are replaced with miter gears to reduce the slipping occurrence. The base enables better stability while the gears are to obtain more accurate readings.	Spring might not operate in error-free conditions. Spring would have high tendency to not return to its original position. This might be due to the large frictional force created from having too many contact points in the design. Smaller gear pitch will cause a potential problem when the gear teeth do not align exactly and cause partial engagement of gears. There are no stoppers to limit rotation of shafts that serve to measure and accumulate the amount of water consumed. Hence, any contact on the shaft will cause variations in readings obtained.

.

While several preliminary ideas were generated at this stage, they were found to lack the defining characteristics of a smart system. Therefore, the research progressed to the next phase, focusing on identifying the root causes of the mechanical limitations present in these initial concepts. Concurrently, further investigation was undertaken to refine and enhance the current device, transforming it into a smart system capable of fostering and improving water-drinking habits among children.

The main problem is identified as the first level in cause-and-effect chain (CEC) analysis. At this stage, the critical limitation addressed was the device’s inability to efficiently record the amount of water consumed. Thus, the primary problem was defined as:

“Initial readings are not saved properly”.

As the CEC diagram is developed, the analysis may point toward human behavior or psychological factors as potential root causes. However, since this project did not encompass these research areas, such causes were intentionally avoided or eliminated. Figure 3 illustrates the CEC diagram constructed for this stage of the research.

The TRIZ approach was employed with reference to the root causes. The same methodology of contradiction formulation and resolution employed in Tan, Ng, Tan, and Lim (2021) [40] was used in this paper. The engineering contradictions were then constructed.

3.1. Sub-Problem 1: Housing of Prototype Is Small

The first engineering contradiction was as follows: if the prototype’s housing is larger, then the spring will remain at its original length, ensuring accurate readings, but the inclusion of excessive mechanical components will increase the device’s weight. Based on this technical contradiction, improving variables (#27 Reliability and #28 Measurement Accuracy) and a worsening variable (#2 Weight of stationary object) were assigned (

Table 3. System parameters for first engineering contradiction.

Engineering Contradiction Variables	System Parameters
Improving Variable	#27 Reliability #28 Measurement Accuracy
Worsening Variable	#2 Weight of Stationary Object

), where each system pair produces a list of inventive principles, as shown in

Table 4. Inventive principles from system parameters.

Pairs of System Parameters	Inventive Principle
#27 and #2	#3 Local Quality #10 Preliminary Action #8 Anti-weight #28 Mechanics Substitution
#28 and #2	#25 Self-service #26 Copying #35 Parameter Changes

. The design objective was to create a smart, portable device adaptable to various environments. Consequently, the inventive principle of mechanics substitution was selected to address this contradiction. This is because converting a mechanical device to a smart and portable device which can adapt to various environments would require the incorporation of sensory means such as optical. The same inventive principle has been used in designing a roof tile transportation and inspection system, where an automated image capture system was used in detecting and inspecting the roof tiles instead of depending on mechanical means requiring operators [50].

The research proceeded by employing another TRIZ technique, which utilizes physical contradictions for problem-solving. This technique, known as the TRIZ Separation Principle, is explained below.

A physical contradiction arises when a single “Control Parameter” (a parameter that can be adjusted) exhibits conflicting values within a system. Essentially, it represents opposing technical requirements for the physical state of an object. There are three primary strategies to resolve a physical contradiction:

• Separation of contradictory requirements;
• Satisfaction of contradictory requirements;
• Bypass of contradictory requirements [62].

Figure 4 displays an illustration to better explain the strategies to resolve physical contradictions. In this study, the physical contradiction was the following: the prototype’s housing must be small to ensure portability, yet large enough to accurately record the amount of water consumed by different users. According to Figure 4, there are four types of separation techniques, which are time, space, condition, and between a part and the whole. Separation in space is a concept to separate opposite requirements in space such as bifocal glasses with both distance and close-up lenses. Separation between conditions is a concept separating the opposing requirements of a condition, which can resolve contradictions in which a helpful process takes place when special conditions exist. For example, ice is solid, but when ice skating, the ice below the skate melts for a fraction of a second, thereby enabling the skates to slide. Separation between a part and the whole is a concept to separate the opposite requirements within a whole object or its parts. If a system must perform contradictory functions or operate under contradictory conditions, try to partition the system and assign one of the contradictory functions or conditions to a sub-system (or several sub-systems). A good example of this concept is the bicycle chain, which has rigid links but is flexible at the system level. Lastly, separation in time is the concept to separate opposite requirements in time. If a system or process must satisfy contradictory requirements, perform contradictory functions, or operate contradictory conditions, try to schedule the system operation in such a way that requirements, functions, or operations that conflict take effect at different times. For example, traffic lights are used to sequence the flow of traffic at different points of time. In this study, the separation in time method was selected as it resolves the contradiction by allowing the characteristic to be larger at one time and smaller at another. It can also address contradictions where a characteristic is present at one time and absent at another. The listed recommended Inventive Principles for Separation in Time are #10 Prior Action, #11 Beforehand Cushioning, #15 Dynamics, #34 Discarding and Recovering, #21 Skipping, and #26 Copying.

Among the listed inventive principles, #26 Copying resolves the contradiction by replacing an object or process with optical copies [64]. Consequently, this principle was selected as the most suitable solution. More information on how the proposed concept utilized this inventive principle is explained in Section 3.2. The same inventive principle has been reported to be utilized in a different field. Despite originally being used in solving problems of physics and chemistry, the increasing impact of software solutions has led to the application of TRIZ in that field. Inventive principle #26 Copying was utilized to perform a shallow copy on data structures instead of a deep copy, thus reducing memory usage and computing time [65].

3.2. Proposed Concept: Smart Water Measuring, Accumulating, and Display Device

Smart systems are defined as advanced technological systems that operate autonomously or collaboratively, integrating functionalities such as sensing, actuation, and control to monitor, analyze, and respond to specific situations. A common definition of smart systems highlights their external, observable functionality and the diverse components necessary for their implementation [66].

Building on the TRIZ methodology and academic resources, the process of identifying root causes led to further modifications, resulting in a refined conceptual idea. This idea was derived from the application of the physical contradictions and separation techniques of TRIZ, which guide the development of new inventions to enhance the device’s effectiveness and robustness. The primary focus was to evolve this concept into a smart device. This section outlines the conceptual idea developed using the 40 Inventive Principles of TRIZ to address the main problem.

The conceptual idea emerged from the integration of several inventive principles, derived from the first root cause identified in the cause-and-effect diagram: the housing was too small. The TRIZ inventive principles applied included mechanics substitution and copying. The smart device comprises several complex electronic components, as illustrated in Figure 5.

At the center of the device is a platform designed to hold the water bottle, enabling the measurement of its volume. A Liquid Crystal Display (LCD) screen displays the amount of water in the bottle. The base plate, which serves as the main structural component, holds all elements in place. To minimize weight, aluminum was selected for both the base plate and the weighing platform, as the device’s operation does not require resistance to heavy force applications.

The inclusion of the LCD screen aligns with the copying principle, as it enables the display of water consumption data for multiple users. This conceptual idea represents a futuristic design, combining smart functionality with the capability to store and manage data for multiple users.

Additionally, the integration of a microcontroller, resulting from the application of the mechanics substitution principle, further enhances the device’s intelligence. The microcontroller allows the device to automatically identify users as they approach it. This functionality is supported by artificial intelligence, advancing the device’s capabilities and positioning it as a smart, cutting-edge solution. The TRIZ flowchart of this study is displayed in Figure 6.

3.3. Component Selection

The smart system integrates various components, including a microcontroller, speaker, weight detection sensor, webcam, and LCD screen. The potential options for each component are outlined in a combination chart, as presented in

Table 5. Components, selection criteria, and types.

Components	Selection Criteria	Types/Choices Available
Microcontroller	Cost Task Efficiency Digital I/O Pins	Arduino Uno Arduino Mega 2560 Arduino Due
Speakers	Size Cost Weight	JBL Logitech Amazon Basics
Weight Detection Sensor (Load Cell)	Range Size Cost	Pneumatic Hydraulic Capacitance
Camera/Webcam	Effective Pixels Frame Rate Compatibility Cost	Logitech B525 Arduino Module Logitech BRIO
LCD Screen Display	Size Resolution Capability (touch) Cost	OEM TFT SPFD 5408 OEM 1602A OEM TFT 3.5inch
Portable Power Supply	Size Capacity Cost Ports Availability	Xiaomi PLM01ZM (Pro) Samsung ULC Yoobao YB-PL12

.

Given that these components had a relatively minor impact on the overall visualization of the smart system, only a screening process was necessary to determine the most suitable options. The screening process for selecting the appropriate components is detailed in the following tables.

In product development, selecting a microcontroller unit requires careful consideration of factors such as maximum processing speed, RAM or ROM capacity, the number and types of input/output (I/O) pins, power consumption, constraints, and development support [67].

Table 6. Screening of microcontrollers.

Selection Criteria	Types of Microcontrollers
	Arduino Uno (Reference)	Arduino Mega 2560	Arduino Due
Cost	0	−	−
Task efficiency (Speed)	0	+	+
Availability	0	0	−
Storage	0	+	0
Digital I/O pins	0	+	+
Sum of +	0	3	2
Sum of 0	5	1	1
Sum of −	0	1	2
Net Score	0	2	0
Rank	2	1	2
Continue?	NO	YES	NO

outlines the screening process used to identify a suitable microcontroller for building the smart system.

Arduino was selected as the microcontroller due to its cost-effectiveness and versatility in performing complex tasks akin to a computer. As an open-source platform, Arduino offers a user-friendly environment for hobbyists, students, and professionals to develop devices that interact with their surroundings using sensors and actuators. It is capable of processing inputs and controlling outputs for a wide range of electronic devices.

Arduino consists of two main components: the hardware, which includes the Arduino development board (the microcontroller unit), and the software, known as the Arduino IDE (Integrated Development Environment), used for writing algorithms and code. The Arduino IDE supports programming in C or C++ [68]. Among the various Arduino boards available, such as the Arduino Uno, Arduino Mega 2560, and Arduino Due, the Arduino Mega was chosen for this project. Its selection was based on its onboard data storage capability, eliminating the need for external memory cards, and its extensive I/O ports, which are essential for building the smart system.

The quality of sound produced by speakers is determined by frequency and amplitude. Frequency defines the pitch, with higher frequencies producing sharper sounds (e.g., a soprano singer’s voice) and lower frequencies generating deeper tones (e.g., a bass guitar or kick drum). A speaker system’s ability to accurately reproduce sound frequencies directly impacts audio clarity.

Table 7. Screening of speakers.

Selection Criteria	Types of Speakers
	Creative Pebble (Reference)	Logitech	JBL
Size	0	+	−
Cost	0	+	−
Weight	0	−	+
Sum of +	0	2	1
Sum of 0	3	0	0
Sum of −	0	1	2
Net Score	0	1	−1
Rank	2	1	3
Continue?	NO	YES	NO

outlines the screening process for selecting a suitable speaker for the smart system.

Logitech speakers were chosen for their affordability and compact design. Their smaller size contributes to the device’s portability, making it easier to transport between locations.

Weight measurement is achieved using a weight sensor, commonly referred to as a load cell. Selecting an appropriate weight detection sensor was a critical aspect of this research.

Table 8. Screening of weight detection sensor.

Selection Criteria	Types of Weight Detection Sensor
	S-Type Beam Load Cell	Load Cell YZC-133 (Reference)	Digital Load Cell HX711
Size	−	0	−
Range	−	0	+
Cost	0	0	0
Availability	−	0	+
Sum of +	1	0	2
Sum of 0	1	3	1
Sum of −	2	0	1
Net Score	−1	0	1
Rank	3	2	1
Continue?	NO	NO	YES

outlines the screening process for choosing a suitable load cell. The selection depends on the required sensitivity and accuracy of the application. For instance, capacitive load cells are the most accurate and sensitive, followed by strain gauge types.

Capacitive load cells are capable of detecting minute changes in force, such as the weight of water, making them ideal for precise applications. In contrast, pneumatic and hydraulic load cells are better suited for less sensitive tasks, such as measuring vehicle weights in car inspection facilities.

A digital load cell was selected through the screening process due to its widespread availability in electronic stores. It is a common add-on component compatible with Arduino microcontroller boards. Additionally, its easy accessibility in local electronics shops ensures quick replacement in case of emergencies, minimizing downtime.

A webcam was also essential for this project, as it captures still images and live video feeds.

Table 9. Screening of camera/webcam.

Selection Criteria	Types of Webcams
	Logitech B525	Arduino OV7670	Logitech BRIO (Reference)
Effective Pixels	+	−	0
Frame Rate	0	0	0
Compatibility	−	+	0
Cost	−	+	0
Sum of +	1	2	0
Sum of 0	1	1	4
Sum of −	2	1	0
Net Score	−1	1	0
Rank	3	1	2
Continue?	NO	YES	NO

outlines the screening process for selecting a suitable webcam for the smart system.

The screening process prioritized compatibility and frame rate, as these features are critical for the smart system’s functionality. The webcam’s frame rate directly impacts its efficiency in detecting individuals in front of the device with minimal human intervention in its operation [69,70]. Additionally, compatibility with the microcontroller and cost were key considerations, particularly for potential replacements in case of damage during operation. Based on these criteria, the Arduino OV7670 was selected as the most suitable webcam.

Table 10. Screening of LCD screen.

Selection Criteria	Types of LCD Screen
	OEM TFT SPFD 5408	OEM TFT 3.5inch (Reference)	OEM 1602A
Size	−	0	+
Resolution	0	0	−
Capability (touch)	+	0	0
Cost	−	0	+
Sum of +	1	0	2
Sum of 0	1	4	1
Sum of −	2	0	1
Net Score	−1	0	1
Rank	3	2	1
Continue?	NO	NO	YES

outlines the screening process for selecting an appropriate LCD screen. The OEM1602A was chosen due to its compact size and lower cost compared to the TFT SPFD screen. Since the LCD screen’s primary function is to display values measured by the weighing platform, touch capability was unnecessary. The screen will also feature a simple rating system to evaluate the child’s water intake, providing an interactive element to engage the child using the device.

Table 11. Screening of portable power supply.

Selection Criteria	Portable Power Supply
	Xiaomi PLM01ZM (Pro)	Samsung ULC (Reference)	Yoobao YB-PL12
Size	+	0	−
Capacity	0	0	+
Cost	+	0	+
Ports Availability	0	0	0
Sum of +	2	0	2
Sum of 0	1	4	1
Sum of −	0	0	1
Net Score	2	0	1
Rank	1	3	2
Continue?	YES	NO	NO

outlines the screening process for selecting a suitable portable power supply. The Xiaomi power bank was chosen for the smart system due to its compact size compared to Samsung and Yoobao power banks. Additionally, the Xiaomi PLM01ZM is more cost-effective, making it the most suitable option.

3.4. Programming Flowchart

A programming flowchart was utilized in this research to outline the intended functionality of the smart system. Flowcharts serve as a foundational tool for documenting program logic during application development [71], which made them a critical preliminary step in designing the Smart Water Measuring, Accumulating, and Display Device. Figure 7 illustrates the flowchart of the smart system.

The flowchart is divided into two sections: the first focuses on facial recognition, while the second monitors the amount of water consumed by each child at one-hour intervals. This monitoring phase includes a reminder mechanism in the form of a buzzer activation, creating an interactive element between the smart device and the child.

To ensure effective facial recognition, a database must be prepared and pre-processed before the device can be utilized. Feature extraction is one of the key strategies required for training the system, enabling it to accurately model and recognize user data.

The smart system is designed to identify up to four different users. This functionality is enabled by a camera positioned beside the measuring platform, which captures images of individuals facing the device. Facial recognition algorithms then process these images to identify the user. To achieve this, feature extraction must be applied to both the database and the camera’s input frames.

However, the system’s operation relies on the integration of multiple components, including an LCD screen, speaker, microcontroller, and power source. Facial recognition cannot be performed solely by the camera; it requires collaboration with these components to function effectively. Due to time constraints in developing the physical device, a laptop-based prototype was used to simulate the smart system’s functionality. Similar face recognition using a laptop camera was previously reported by others [72,73].

As such, the smart device is capable of:

• Individual facial recognition: identifying one registered user at a time via laptop-connected camera.
• Persistent data display: showing multiple users’ cumulative consumption on LCD.

Thus, experimental procedures were designed based on these limitations.

3.5. Experimental Procedures

Prior to conducting the experiments outlined in this section, research ethics approval (approval number EA0352022) was obtained from the Research Ethics Committee of Multimedia University. Additionally, written consent was secured from the parents of the participating children.

The first experiment aimed to monitor the child’s water consumption, serving as a key method to evaluate whether the device effectively cultivates a water-drinking habit. The experiment was conducted over a fixed period, with a maximum duration of four hours.

Testing was carried out in two distinct environments:

Situation A: A controlled environment without the smart device.

Situation B: An environment incorporating the smart device.

Before conducting the second test (Situation B), the child was instructed on how to use the device. This ensured familiarity with its functionality prior to evaluation.

Null Hypothesis, Ho	:	There is no significant difference in having the smart system between Situation A and Situation B.
Alternative Hypothesis, Ha	:	There is a significant difference in having the smart system between Situation A and Situation B.
Fixed	:	The type, size, and design of the container (or cup) used by the child throughout the experiment.
Recorded value	:	Volume of water drunk (mL).
Conditions	:	The experiment will be carried out in two stages, which are Situation A and Situation B, respectively. Situation A (the control) should be tested before Situation B in order to avoid bias in the data collected. Each experimental stage will be carried out at the same time of day on two consecutive days (in order to rule out the effects of surrounding temperature). The guardian/teacher does not need to instruct the child to drink water for either Situation A or Situation B. A water jar is to be placed on the device and a continuous flow of water is ensured at all times (by filling the jar to the brim once empty). The child will use their own containers to consume water. Child should not have consumed water or liquid for at least one hour before the experiment commences.

Steps:

• Firstly, the jar will be filled to the brim and the amount of water is recorded. Then, the jar is placed onto a table and a stopwatch is started to record the time.
• At every one-hour interval, parents are to remind their child to drink water.
• At the end of the fourth hour, the stopwatch is stopped and parents are to calculate the total amount of water drunk by the child. The result obtained will be recorded in a table.
• The experiment will be repeated for the second stage on a different day where instead of placing the filled water jar on the table, the jar is placed onto the device and the total amount of water will be measured automatically by the device.
• Then, parents are not to remind their children at every hour interval as they did in the first phase of the experiment. However, parents are to monitor the condition of the jar and ensure sufficient water is in the jar.
• At the end of the experiment (after four hours), parents are to request the child to stand in front of the device for a short time. Then, the smart system will display the total amount of water drunk by the child on the LCD screen.
• The result obtained is then tabulated on the same table (in a separate column).

The second experiment focused on measuring the time taken for a child to consume a specific volume of water within a day, based on the recommended intake for their age group. This experiment also allowed the child to engage with the educational features of the device.

Similar to the first experiment, testing was conducted in two environments:

Situation A: A controlled environment without the smart device.

Situation B: An environment incorporating the smart device.

Null Hypothesis, Ho	:	There is no significant difference in having the smart system between Situation A and Situation B.
Alternative Hypothesis, Ha	:	There is a significant difference in having the smart system between Situation A and Situation B.
Fixed	:	The type, size, and design of the container (or cup) used by the child throughout the experiment.
Recorded value	:	Time taken in hours to finish up the prescribed amount of water for each child.
Conditions	:	The experiment will be carried out in two stages (Situation A and Situation B). Situation A (the control) should be tested before Situation B in order to avoid bias in the data collected. Each experimental stage will be carried out at the same time of day on two consecutive days (in order to rule out the effects of surrounding temperature). The child will use their own containers to consume water. The child would need to be taught how to use the device. The guardian needs to be the one that instructs the child to drink water in Situation A, and not the experimenter. However, the guardian can only instruct once, which is at the beginning of the experiment where he/she informs the child to remember to consume water. The guardian does not need to instruct the child to drink water in Situation B. Child should not have consumed water or liquid for at least one hour before the experiment commences. The age range data should correspond with the range of the prescribed amount of water consumption.

Steps:

• For the first stage, the jar is filled with a specific amount of water depending on the prescribed amount of water consumption for the child.
• Then, the jar is placed onto a table and a stopwatch is started.
• Once the water in the jar has been fully consumed, the stopwatch is stopped.
• The time taken for the child to drink the water in Situation A is then recorded.
• The second phase of the experiment is carried out by filling the jar with the same amount of water again. However, the jar is placed onto the device instead and a stopwatch is used to start recording the time once the jar is placed onto the device.
• The stopwatch is then immediately stopped once the water has been fully consumed.
• The time taken to empty the jar for Situation A and Situation B are recorded and tabulated in a table.

The third experiment observed the frequency of water consumption within a fixed period (maximum five hours). This method indirectly assessed the child’s interest in using the smart device. The frequency of drinking served as an indicator of whether the habit of drinking water had been cultivated. The experiment included two conditions:

Situation A: A controlled environment without the smart device.

Situation B: An environment incorporating the smart device.

Null Hypothesis, Ho	:	There is no significant difference in having the smart system between Situation A and Situation B.
Alternative Hypothesis, Ha	:	There is a significant difference in having the smart system between Situation A and Situation B.
Fixed	:	The type, size, and design of the container (or cup) used by the child throughout the experiment.
Recorded value	:	Number of times the child drinks water within a five-hour period.
Conditions	:	The experiment will be carried out in two stages (Situation A and Situation B). Situation A (the control) should be tested before Situation B in order to avoid bias in the data collected. Each experimental stage will be carried out at the same time of day on two consecutive days (in order to rule out the effects of surrounding temperature). The child will use their own containers/cup to consume water. The child would need to be taught how to use the device. Child should not have consumed water or liquid for at least one hour before the experiment commences.

Steps:

• The jar of water is filled at a full level and placed on a table.
• Then, the stopwatch is started using a five-hour-period timer.
• The child is then observed within the period and whenever he/she drinks water from the jar, the count (frequency) is recorded.
• The number of times the child drinks water within the five-hour period is then recorded and tabulated in a table.
• The experiment is repeated again with the jar of water filled at a full level.
• However, this time the jar is placed on the measuring platform of the smart device.

The frequency (times that the child approaches the device to drink water) is then tabulated and recorded as well.

4. Discussion

Figure 8 and Figure 9 illustrate the front and top views of the prototype for the Smart Water Measuring, Accumulating, and Display Device. Figure 8 was captured before conducting the experiments in the designated classroom with the children. As observed, each child used a water bottle of a similar size, ensuring consistency in the experimental setup.

Table 12. Details of device prototype.

Device name	Smart Water Measuring, Accumulating, and Display Device
Function	Measure, accumulate, and display water
Dimensions	450 (W) × 350 (L) × 340 (H) mm
Weight	4.0 kg
Materials	Aluminum, acrylic, and Polyethylene Terephthalate

shows a list of details of the prototype of the smart device.

4.1. Data Obtained from Experiments with Prototype

Table 13. Amount of water consumed by the child.

Child #	Without Device (mL)	With Device (mL)
1	415	523
2	512	525
3	290	519
4	312	510
5	350	493
6	180	431
7	303	502
8	120	230
9	121	432
10	150	395
11	230	518
12	208	430
13	170	350
14	190	353
15	210	460
Average (mL)	250.7	444.7

,

Table 14. Frequency of drinking water.

Child #	Without Device	With Device
1	4	9
2	6	8
3	5	7
4	3	6
5	3	6
6	2	6
7	3	9
8	2	6
9	1	4
10	2	4
11	2	5
12	4	7
13	3	8
14	2	7
15	7	9
Average	3 times	6 times

and

Table 15. Time taken to consume fixed volume of water.

Child #	Without Device (min)	With Device (min)
1	135	115
2	155	100
3	180	135
4	115	110
5	108	75
6	150	110
7	120	117
8	210	180
9	240	185
10	210	145
11	135	120
12	240	180
13	200	120
14	230	140
15	120	75
Average (mins)	169.9	127.1

present the experimental results obtained from each of the tests conducted on the prototype. Prior to the experiments, the children’s water consumption was monitored without the smart device. A digital weighing scale and stopwatch were used to manually collect baseline data, which was later compared to results obtained using the smart device. This approach was implemented to mitigate the Hawthorne effect, which could otherwise influence the outcomes when testing with the smart device [74].

The experiment began with registering each child on the smart device, a process that took approximately 30 min. A timer was then set to record data over a four-hour period. Each time a child approached and used the device, it recorded the amount of water consumed. These results were subsequently compared to those from the control condition, where the smart device was absent.

From Table 13, it is evident that the amount of water consumed by each child increased with the use of the smart device. The second experiment, which measured the frequency of water consumption within a fixed period, was conducted concurrently with the first experiment.

As shown in Table 14, the presence of the smart device appeared to motivate children to drink water more frequently. When used alongside peers and classmates, the device yielded positive results, fostering greater engagement and improved hydration habits among the test group.

The final experiment measured the time taken to consume a fixed amount of water (see Table 15). This test was conducted on a separate day, as the first two experiments collectively occupied half a day. Additionally, all experiments were repeated at the same time of day to control for potential variations in surrounding temperature that could influence the results.

The recommended daily water intake for children aged 6 to 12 is 1.6 L. However, for this experiment, a fixed volume of 800 mL was set to ensure completion within the limited timeframe of school hours.

At the conclusion of the experiment, it was observed that most children consumed the fixed volume of water in a shorter time compared to the control condition. Although improvements were noted across all three experiments, the data required validation to determine whether the differences between using and not using the smart device were statistically significant. Therefore, a detailed statistical analysis is presented in the following sub-chapter.

4.2. T-Test Analysis

The t-test conducted in this research was a two-sample t-test, which compares the means of two related groups—such as the same group tested before and after an intervention. In this study, the t-test was used to evaluate the differences in results obtained without and with the smart device. The first step in performing the t-test involved defining the null hypothesis (H₀) and the alternative hypothesis (H₁).

The null hypothesis (H₀) stated that there is no significant difference in outcomes when using the smart device, while the alternative hypothesis (H₁) posited that there is a significant difference. To ensure the validity of the t-test, the following conditions were met:

• Data for each group were obtained through random sampling.
• The variances of the two groups were unequal.
• Data in each group were normally distributed.
• Measurement values were continuous.

After collecting the experimental data, a normality test was conducted using JMP version 16, a statistical analysis software. Figure 10 displays the normality test results for each experiment. As observed, all experimental data fell within a normal distribution curve, confirming that the data could be analyzed using a t-test to determine whether there was a significant difference between using and not using the smart device. The next step involved calculating the p-values for each experiment. Using JMP software, the p-values were computed and organized in a table, as presented in the following sub-section.

4.3. Tabulation of Significance Levels

Using JMP software, the p-values were calculated and tabulated for each experiment, as shown in

Table 16. Results for each experiment.

No	Experiment	Sample Size, N	t-Value	p-Value
1	Amount of Water Drunk	15	−5.34	0.000
2	Frequency of Drinking Water	15	−5.69	0.000
3	Time Taken to Consume Water	15	2.79	0.010

. For a normally distributed dataset, a significance level of α = 0.05 was established prior to computing the p-values. The results of the p-value calculations for each experiment are presented in the table below.

The p-values for the amount of water consumed, frequency of drinking water, and time taken to consume water were all less than the significance level of α = 0.05. This indicated a 95% confidence level to reject the null hypothesis (H₀), which stated that there is no significant difference when using the smart device, and to accept the alternative hypothesis (H₁), which asserted that there is a significant difference when using the smart device.

The overall results from each experiment are summarized in Figure 11. Improvements were observed in the amount of water consumed, the frequency of drinking water, and a reduction in the time taken to consume a fixed volume of water when the smart device was utilized.

4.4. Discussions and Findings

One of the primary objectives of this research was to evaluate the effectiveness of the smart device in encouraging children to drink more water. However, certain experimental variables did not fully represent the device’s intended commercial use. For instance, the smart device was designed for household settings, but the experiments were conducted with a group of preschool children rather than individual families.

This approach was adopted due to time constraints and the need for a larger sample size. Parents were excluded as participants to minimize external influences, such as reminders to drink water, which could skew the results. The experiments spanned three days, beginning without the smart device to eliminate the Hawthorne effect [74], followed by testing with the device introduced.

Several key observations emerged during the experiments, leading to the following major findings:

• Participants frequently reminded each other to drink water.
• Cognitive performance among participants improved.
• Participants became more aware of the amount of water they consumed.

Since the smart device displayed each participant’s water intake, it was observed that children compared their consumption and encouraged one another to drink more water throughout the experiment. This finding highlights the device’s potential to foster positive water-drinking habits, as peer reminders and encouragement are critical components of habit formation [75]. By integrating the smart device into daily routines, it can serve as a tool to encourage interaction not only among peers but also within families. This collaborative engagement can further support the cultivation of healthy water-drinking habits [76].

Another significant finding was the improvement in cognitive performance observed among participants during the experiment. Teachers noted measurable increases in classroom engagement, including sustained attention during lessons, more frequent note-taking, and active participation, all of which are patterns that align with established research on hydration and cognitive function in children. This observation aligns with previous research, which establishes a correlation between water consumption and cognitive performance [77,78]. Additionally, during the experiment with the smart device, participants demonstrated increased awareness of their water consumption. The device’s continuous display of total water intake allowed children to easily monitor and track their progress. This contrasted with the condition without the smart device, where no visual record of water consumption was available.

By incorporating the smart device into the classroom setting, it not only provided real-time feedback on individual water intake but also created opportunities for children to learn by observing their peers’ consumption patterns [79]. This observational learning, also referred to as social learning or modelling, is a form of learning that involves observing and imitating the behaviors of others.

5. Conclusions

Observing the results obtained from all experiments, the Smart Water Measuring, Accumulating, and Display Device was successful in encouraging children to drink more water. The smart device acted as a tool that assists in cultivating a good water-drinking habit among children. The results obtained from the experiments carried out with the prototype led to the conclusion that the smart device had successfully achieved the objectives. This was because there was a significant difference in the results obtained in all experiments. However, a longer period of monitoring is still required to confirm that the habit has been cultivated.

5.1. Participants Reminded Each Other to Drink Water

Besides the experimental results obtained, there were a few significant changes that were observed along the way during the final rounds of the experiment as well. One of the observations was that participants would remind one another to drink more water. Since the amount of water drunk by all participants was clearly displayed on the device, it was observed that participants were comparing and reminding each other to continue drinking water. This significant interaction has a positive impact because one of the important elements of habit formation among children is encouraging and reminding [75]. That is why parents are said to have an important role in training children to have good water-drinking habits from a young age [76]. As observed from the experiments, the Smart Water Measuring, Accumulating, and Display Device could be an instrument that encourages interaction between members of a family in the field of cultivating good water-drinking habits.

5.2. Improved Cognitive Performance Among Participants

Another major finding that was discovered was an improved cognitive performance among participants throughout the lesson period. This finding correlates with past findings whereby cognitive performance is influenced by water consumption [9,77]. Participants were observed taking down notes more often and tending to be more attentive during lesson periods whenever the teacher was explaining in front of the classroom. This finding supports and proves that drinking more water will result in an improvement in the cognitive performance of a child [78].

5.3. Participants Aware of Amount of Water Consumed

Another significant observation was that participants were aware of the amount of water consumed throughout the day. While adapting to the usage of the smart device, the participants know the current amount of water they have consumed at that particular period of time. This was made possible with the total amount of water constantly displayed on the screen. It was different compared to the condition without the smart device, where there was no form of display of the amount of water drunk by the participant. Research by Kaushik (2007) stated that children consumed less water when the availability of water was controlled or prohibited [80]. The statement “out of sight, out of mind” accurately describes the condition of how, when the child does not see, they forget to drink. Hence, the smart device also acts as a visual reminder for the child throughout the day, which encourages the child to drink more water.

5.4. Challenges and Limitations of Study

In pursuing any research, acknowledging the limitations and challenges is crucial for a comprehensive understanding of the study’s scope and implications. These factors could impact the reliability and validity of the findings. Limitations in general are the inherent constraints or weakness in the design, methodology, or data which may affect the overall results. Challenges basically are obstacles encountered during the research process that hinder the progress of the study. This section will discuss a few primary limitations and challenges that were faced in this research.

One of the limitations of this research was that the prototype was tested within a fixed group of children aged 6 years old. This was due to the time constraint and the need to obtain a significant number of participants to participate in the experiments. However, this was also done in order to rule out factors such as age that could affect the amount of water drunk by the participants during the test, which could have resulted in data with a not-normal distribution curve. Another limitation was the lack of a model being used in measuring the improvement in cognitive performance. It would benefit this study if a suitable model or more appropriate test were used to measure the improvement, while providing quantitative data for comparison with other studies.

Additionally, since the participants of this study were only children, we could not exactly measure the parent–child interaction and its impact on the children’s water intake. However, the interactions among children while using the device may serve as a useful indicator of social influence on drinking behavior.

Lastly, since this study was conducted over a short duration, it was difficult to determine whether the observed increase in water intake was sustained or simply a short-term response due to the children’s curiosity.

A few major challenges that were observed throughout the research were the ability to carry out the tests and experiments in the homes of different families, obtaining consent from pre-school management, and time consumption in performing tests and experiments.

5.5. Directions for Future Research

With the advancement of technology and Artificial Intelligence (AI) in our daily lives, the future of the Smart Water Measuring, Accumulating, and Display Device also lies in technological improvements. The functions of the smart device are limited by location constraints. If the smart device is placed in the child’s home, he or she will not be able to fully utilize the smart device. The child would need to carry along the smart device, which is physically operated by placing the water bottle onto the measuring platform. These constraints can be overcome if the size of the smart device could be designed to be smaller. An additional feature that would be helpful is an individual wireless sensor placed below the bottom of the child’s water bottle. This would enable the child to move around freely and perform daily activities without having to return to the smart device repeatedly.

The topic of children not drinking enough water and staying well-hydrated has been a long-studied topic because of the severe impact that it brings. The smart device produced in this research might not be the final solution to the challenge of cultivating good water-drinking habits among children. However, its development could direct the community one step closer to finding the best possible solution. The invention of this smart device introduces a creative solution from a new and different approach. But more research is needed to verify its effectiveness in promoting and forming a good habit. Since it can be an educational tool, research should also consider the interaction it creates in the context of a family in the near future.